Jump to content
Welcome to our new Citrix community!
  • 0

CH8.1 GFS2 attach error, "The SR failed to complete the operation" is already mounted or busy


Sergey Polyakov

Question

One of GFS2 SRs begin work incorect. One by one hosts mark this SR as disconected. After reboot all hosts in pool, SR stil have a problem.

 

 xe pbd-plug uuid=3fab3c98-f88e-23e5-ca50-a3d800741659
There was an SR backend failure.
status: InternalError
stdout: /usr/bin/mount -t gfs2 -o noatime,nodiratime /dev/mapper/360060e8012b276005040b27600002101 /var/run/sr-mount/614e53db-2499-6f73-9bd2-500a473b8ee4 exitted with non-zero code 32: mount: /dev/mapper/360060e8012b276005040b27600002101 is already mounted or /run/sr-mount/614e53db-2499-6f73-9bd2-500a473b8ee4 busy

stderr: <unknown>

 

if try repear :

 fsck.gfs2 -y /dev/mapper/360060e8012b276005040b27600002101
Initializing fsck
Segmentation fault (core dumped)

 

Link to comment

18 answers to this question

Recommended Posts

  • 0
1 hour ago, Tobias Kreidl said:

You see this on all your hosts? Might be corrupted if you cannot even get fsck to run on any of them. However, I'd look for the /dev/xyz device and run fsck on it instead of on the mamper link. You want to run fsck on the raw device.

 

-=Tobias

All hosts.

 

360060e8012b276005040b27600002101 dm-18 HITACHI ,OPEN-V          
size=4.0T features='0' hwhandler='0' wp=rw
`-+- policy='service-time 0' prio=1 status=active
  |- 0:0:0:0  sda  8:0    active ready running
  `- 1:0:1:0  sdbg 67:160 active ready running

 

fsck.gfs2 -vy /dev/dm-18
Initializing fsck
Segmentation fault (core dumped)
 

Link to comment
  • 0

That does not sound good. Can you try fsck on a different superblock (add "-b 32768" to the fsck command)?  I'd start by locating any lock files sand deleting them. Google "Initializing fsck Segmentation fault (core dumped)" to see what others have experienced and a couple of suggestions. Have you tried to power cycle the storage unit, itself, not that this would necessarily help, but there may be something hung on that end? Do you see any further clues in /var/log/SMlog on your hosts?

Link to comment
  • 0
14 minutes ago, Tobias Kreidl said:

That does not sound good. Can you try fsck on a different superblock (add "-b 32768" to the fsck command)?  I'd start by locating any lock files sand deleting them. Google "Initializing fsck Segmentation fault (core dumped)" to see what others have experienced and a couple of suggestions. Have you tried to power cycle the storage unit, itself, not that this would necessarily help, but there may be something hung on that end? Do you see any further clues in /var/log/SMlog on your hosts?

 

fsck -b 32768  /dev/dm-18
fsck from util-linux 2.23.2
fsck.gfs2: invalid option -- 'b'
Please use '-h' for help. 

Link to comment
  • 0
27 minutes ago, Tobias Kreidl said:

Do a "man fsck.gfs2" to see what syntax they require to specify an alternate block. Might not be an option in fsck.gfs2 or it may automatically use an alternative block... am frankly not quite sure.

 

"man fsck.gfs2" doesn't help


 But I found the host that most likely locks the disk

Only one host has a busy error.

fsck.gfs2 -y /dev/dm-17
Initializing fsck
Device /dev/dm-17 is busy.

 

Rebooting the host didn't help. Disk is still busy

 

Link to comment
  • 0
19 hours ago, Tobias Kreidl said:

As mentioned before, you may need to identify and clear lock files. It seems to think the volume is busy, which may indicate a stale lock.

Didn't find anything that could use this drive.

multipath -F
does not show that this disc is in use.
I think the problem is in the logic of the file system GFS2
I shut down all the pool servers at the same time.
This did not solve the problem

Link to comment
  • 0
13 hours ago, Tobias Kreidl said:

 

I don't know why, but I see this Disk by command "fdsik -l"

 

Disk /dev/mapper/360060e8012b276005040b27600002101: 4398.0 GB, 4398046511104 bytes, 8589934592 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk label type: gpt
Disk identifier: E2268763-A110-4BB4-AC22-EE35F00BC505


#         Start          End    Size  Type            Name
 1     46139392     83888127     18G  Microsoft basic 
 2      8390656     46139391     18G  Microsoft basic 
 3     87033856   8589934558      4T  Linux LVM       
 4     83888128     84936703    512M  BIOS boot       
 5         2048      8390655      4G  Microsoft basic 
 6     84936704     87033855      1G  Linux swap    

 

And I don't see any other GFS2 drives

Link to comment
  • 0

that sounds like a hardware failure then. Are you using HBA (Fibre Channel) storage or iSCSI? fdisk -l should show the disk if it's visible to the host.

 

It's also a bit odd that sda appears in your multipath, or are you on a system that is booted from SAN?

 

/dev/mapper/360060e8012b276005040b27600002101 is your system disk and not an SR disk for GFS2.

Link to comment
  • 0

Hi.

The problem was caused by a mistake action, the wrong drive was selected during OS installation.

Tried using photorec_static and testdisk_static, doesn't help

Deleted all paritions by fdisk, after that I was able to run fsck.gfs2

 

Now LS shows me some files

but when I try to mount sr I get the error 

 

xe pbd-plug uuid=1604057d-50f4-f2f2-e7db-f34860b7dc79 
There was an SR backend failure.
status: AttributeError
stdout: 'NoneType' object has no attribute 'lun_size'
stderr: <unknown>

 

 

 

 

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...